fix call to tri-diagonal solver in MOM_wave_structure.

[raphaeldussin]

The call to the solve_diag_dominant_tridiag had arguments mixed-up:

subroutine solve_diag_dominant_tridiag( Al, Ac, Au, R, X, N )

lower diagonal: c_diag = Al in the function prototype
midlle diagonal: b_diag = Ad NOT in the function prototype
upper diagonal: a_diag = Au in the function prototype

the function call for Ac of the middle diagonal, defined as

Ad = Ac+Al+Au

Ac=Ad-Al-Au

so that the correct argument for Ac is b_diag - a_diag - c_diag = b_diag - (a_diag + c_diag)

[codecov]

Codecov Report

Merging #154 (335a5d9) into dev/gfdl (c9dd804) will not change coverage.
The diff coverage is n/a.

❗ Current head 335a5d9 differs from pull request most recent head 6940115. Consider uploading reports for the commit 6940115 to get more accurate results

@@            Coverage Diff            @@
##           dev/gfdl     #154   +/-   ##
=========================================
  Coverage     34.02%   34.02%           
=========================================
  Files           259      259           
  Lines         70204    70204           
  Branches      13010    13013    +3     
=========================================
  Hits          23884    23884           
  Misses        41822    41822           
  Partials       4498     4498           

Impacted Files	Coverage Δ
src/diagnostics/MOM_wave_structure.F90	0.00% <ø> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c9dd804...6940115. Read the comment docs.

[Hallberg-NOAA]

I agree that this change corrects an error in the arguments to the (altered) tridiagonal solver that was introduced with
mom-ocean@7062289 .

The altered code is technically correct as written. However, the MOM6 style guide at https://github.com/NOAA-GFDL/MOM6/wiki/Code-style-guide#array-syntax does not encourage the use array-syntax math in subroutine arguments. Given that b_diag is not used anywhere else, apart from in this subroutine call, perhaps it would make sense to define a new variable that is this argument, and to replace the expressions setting b_diag (or perhaps leave the expressions for b_diag in comments for clarity).

[marshallward]

Gaea regression: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/-/pipelines/16102 ✔️

It was decided that the issues around array syntax are no worse than the existing form, and can be addressed in a later PR.